copyright in music: a role for the principles of reverse engineering

Copyright in music: a role for theprinciples of reverse engineeringlest_175 653..673

James GriffinLecturer, University of Exeter

The rise of popular music in the twentieth century has raised questions about the appro-priateness of the current system of copyright law. Copyright law is based around the notionof the individual ‘romantic’ author, an individual who creates with his own innatethoughts. Copyright law provides an exploitable property right to authors – a right, in rem,which may be exercised against the rest of the world. It is a right that may be sold andtransferred, a right to which fiscal value may be placed. The property paradigm ofcopyright is one that is exclusionary. Popular music reveals that copyright works may becollaborative in nature, and this can bring into question whether an exclusionaryproperty-based model is appropriate. Historically, copyright has not always been basedaround the property paradigm; some early cases highlighted the ‘merit’ of the potentiallyinfringing work, and they focused on the manner of creation of that potentially infringingwork. Some later cases have also emphasised the manner of creation of a copyright work.These are cases that concern what is termed ‘reverse engineering’ – a modern term thatencapsulates how an earlier work is used in a later work. Paradigmatically, to focus onreverse engineering is to mark a move away from the property paradigm of copyright. Thispaper argues that to institute such a methodological approach would lead to a moreaccurate ontology and would thus lead to more efficient legal regulation.

INTRODUCTION

In May 2005, Lionel Sawkins1 won his case against Hyperion Records.2 Sawkins wasa musicologist and he had prepared performing editions of works by baroque com-poser Lalande. Hyperion Records had used four of Sawkins’ performing editionswhen they made and sold sound recordings on compact disc. Mummery LJ, in theCourt of Appeal,3 found that copyright subsisted in the performing editions and thatSawkins’ copyright had been infringed.

The case was met with controversy. For instance, Rahmatian4 argued that theinterests of small recording studios such as Hyperion Records were not sufficientlytaken into account. There were arguments that the approach of Mummery LJ was

1. For details see the website available at http://www.lionelsawkins.co.uk/Export2.htm.2. Sawkins v Hyperion [2005] EWCA Civ 565, [2005] 1 WLR 3281.3. Ibid, at 3284–3298.4. A Rahmatian ‘The concepts of “musical work” and “originality” in UK copyright law –Sawkins v Hyperion as a test case’ (2009) 40 International Review of Intellectual Property andCompetition Law 560 at 561 citing M Perlman How a French Baroque Motet is like aMelanesian Folk Song: An Ethnomusicologist Considers the Sawkins v Hyperion Case, avail-able at http://www.andante.com/article/article.cfm?id=25873. See also the blog entry availableat http://www.overgrownpath.com/2005/05/hyperion-records-face-catastrophic.html and http://www.therestisnoise.com/2005/05/the_pierre_mena.html.

Legal Studies, Vol. 30 No. 4, December 2010, pp. 653–673DOI: 10.1111/j.1748-121X.2010.00175.x

© 2010 The Author. Legal Studies © 2010 The Society of Legal Scholars. Published by Blackwell Publishing, 9600Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA

expanding copyright in new and undesirable ways.5 Sawkins’ case is an excellentelucidation of how copyright law has become obsessed with constructing a propertyparadigm of the ‘work’. The method and manner by which the copyright work wascreated and distributed was overlooked in the desire to propertise and assess the fiscalvalue of a work.

Copyright, in revolving around the property paradigm, has adopted a methodologi-cal ontology which no longer bears sufficient resemblance to the underlying content.The rise of remixing and sampling, both quantitatively and qualitatively, has meantthat the notion of intellectual works as property is no longer ontologically appropriate.The law could be reformed, to move away from the property paradigm and embracea paradigm that fully considers the notion of remixing. This could be a system whichfully explores the intricacies and details of re-use, and this paper argues that a nascentsystem of that type can be found in cases dealing with reverse engineering.

AN INTRODUCTION TO THE BASIC PRINCIPLES OFCOPYRIGHT IN MUSIC

Copyright is defined as a ‘property right’ in s 1(1) of the Copyright, Designs andPatent Act (CDPA) 1988:

‘Copyright is a property right which subsists in accordance with this Part inthe following descriptions of work –

(a) original literary, dramatic, musical or artistic works,(b) sound recordings, films [or broadcasts] and(c) the typographical arrangement of published editions.’6

Copyright in relation to a piece of music may involve copyright over the musicalwork itself and, in addition to this, there can be copyright over the lyrics. There mayalso be a copyright over the sound recording. These works should be fixed at somestage7 and have sufficient originality (though a sound recording does not need to beoriginal).8 A ‘musical work’ is a work ‘consisting of music, exclusive of any words oraction intended to be sung, spoken or performed with the music’.9 It is not furtherdefined, but following Sawkins, a broad definition may be supposed:

5. E Vulliamy ‘£1m legal bill rocks a musical institution’ The Guardian 23 December 2005,available at http://www.guardian.co.uk/uk/2005/dec/23/arts.artsnews. Note also the followingtaken from Hyperion Records website: ‘The collateral damage caused by this decision not onlywill affect the prosperity of the company but also the dozens of artists and groups, producers,engineers, composers, music publishers, and musical editors but most importantly the record-buying public whose access to rare and collectable repertoire served by Hyperion, and perhapsmany of the other record labels, will be severely diminished’; from A Robinson ‘HyperionRecords Ltd v Dr Lionel Sawkins: it’s like that and that’s the way it is’ (2005) 16 EntertainmentLaw Review 191 at 191.6. Copyright, Designs and Patents Act 1988 (CDPA 1988), s 1(1).7. G Davies, K Garnett and K Harbottle (eds) Copinger and Skone James on Copyright(London: Sweet and Maxwell, 15th edn, 2005) pp 3–47, and the comments of Mummery LJ inSawkins v Hyperion above n 2, at 3295.8. CDPA 1988, ss 1(1) and 3(2).9. Ibid, s 3(1).

654 Legal Studies, Vol. 30 No. 4

© 2010 The AuthorLegal Studies © 2010 The Society of Legal Scholars

‘In principle, there is no reason for regarding the actual notes of music as theonly matter covered by musical copyright, any more than, in the case of a dramaticwork, only the words to be spoken by the actors are covered by dramatic copyright.Added stage directions may affect the performance of the play on the stage or onthe screen and have an impact on the performance seen by the audience. Stagedirections are as much part of a dramatic work as plot, character and dialogue.’10

Mummery LJ, following a long line of authority,11 emphasised the labour, skill andeffort of the author.12 He stated that performing indicators, tempo and performancepractice indicators could all be capable of obtaining copyright protection, due to thelow threshold of originality. The subsistence of copyright has served to carve up piecesof music into various component proprietary parts and, as Horn has identified,13 thishas come to mean that some of the copyrights works are more valuable than others.For instance, musicians will prefer to gain certain copyrights in order to gain moreroyalties, as in the case of Hayes,14 where vocalist Aniff Akinola15 sought a 10% shareover several copyrights in the musical works rather than confining his royalties tocopyrights in the lyrics.

Once the copyrights over aspects of a piece of music have been identified andestablished, the copyright in question may take many forms. There are, for instance,rights over reproduction, distribution and adaptation.16 In relation to the reproductionright for a musical work, there has to be derivation from the original work and thetaking of a substantial part.17 In relation to the latter part of the test, for music it hasbeen held that the court should focus on ‘aural similarity’.18 It will be argued in detailimmediately below that the tests utilised for infringement sideline the nature of thecreative process involved in musical works and, again, emphasise proprietary controlover copyright works. By parcelling up pieces of music into proprietary parcels, thereis insufficient regard to the impact on future music creation.

In sum, the current approach of the law is to discourage content recipients andmusical authors in re-using certain elements of copyright works. This is not princi-pally due to a nuanced balancing exercise, but the proprietary notions that copyrighthas proposed. The paper will now investigate how that property paradigm has becomeestablished; how the development of music runs counter to that property paradigm incase-law; and, finally, how a new approach could be established which moves awayfrom the property paradigm and more accurately reflects the process of makingmusical works.

10. Sawkins v Hyperion, above n 2, at 3295.11. N Gravells ‘Authorship and originality: the persistent influence of Walter v Lane’ [2007]Intellectual Property Quarterly 267.12. Sawkins v Hyperion, above n 2, at 3295.13. D Horn ‘Some thoughts on the work in popular music’ in M Talbot (ed) The MusicalWork: Reality or Invention (Liverpool: Liverpool University Press, 2000) pp 14–15.14. Peter Hayes v Phonogram Ltd [2003] ECDR (11) 110.15. See the website available at http://www.soundonsound.com/sos/feb05/articles/studiosos.htm.16. CDPA 1988, s 16.17. Designers Guild v Williams [2000] 1 WLR 2416.18. Williamson v Pearson [1987] FSR 97; Sawkins v Hyperion, above n 2, at 3295.

Copyright in music 655


COPYRIGHT AS A PROPERTY RIGHT

Copyright came to be developed as a property right because of historical circum-stance. Three hundred years ago, at the time of the Statute of Anne,19 it was commonpractice for authors to assign their copyrights to publishers, just as they would assignownership of any other chattel.20 Concurrently, publishers would use the names ofauthors as a form of categorisation, to aid the public in deciding which types of worksthey would like to buy.21 A similar development occurred in relation to music, namelythat those behind the distribution of music used the names of composers to makeidentification easier to achieve.22 Distributors then utilised proprietary notions tocontrol dissemination and performance of works such as music.

Today, the consequence of this is clear. As Horn23 noted, the music industry oftenuses the term ‘work’ in relation to music, reflecting the wording of the CDPA 1988 thatmusic is a copyright work.24 However, those who produce the music do not tend to usesuch a term – the reality is that music is comprised of a number of earlier pieces, andthey tend to treat music as more fluid than a set object.25 The phrase ‘work’ denotesspecific boundaries, boundaries which are not present as viewed from the vantage ofthe musician.

In addition, recording studios and right holders prefer clearance on the re-use ofany of their own copyright works.26 Permitting re-use could reduce professionalvaluation of their proprietary interests. This is particularly important when drawing upcontracts, when there needs to be the division of profits.27 The reality of the musicindustry today is that there are many bodies dependent on the division of profits:distributors, record labels, managers, radio stations, collecting societies and so forth.28

The music industry itself sees ‘property’ as a critical tool in the way in which it runsits business. Without the property concept, it argues, how can it exploit music?29

However, there are many instances where the music industry has utilised the propertyconcept to its considerable disadvantage. For instance, the insistence on focusing onindividual musicians meant that the industry failed to foretell the coming of age of

19. Statute of Anne 1710, 8 Anne Ch 19, the long title of which is ‘An Act for the Encour-agement of Learning, by vesting the Copies of Printed Books in the Authors or purchasers ofsuch Copies, during the Times therein mentioned’(the Statute of Anne).20. B Kaplan An Unhurried View of Copyright (New York: Columbia University Press, 1967)pp 8–9: ‘I think it nearer the truth to say that publishers saw the tactical advantage of puttingforward authors’ interests together with their own, and this tactic produced some effect on thetone of the statute’.21. W Straw ‘Authorship’ in B Horner and T Swiss Key Terms in Poplar Music and Culture(Malden, MA: Wiley-Blackwell, 1999) p 200.22. D Brackett ‘Music’ in Horner and Swiss, ibid, p 129.23. Horn, above n 13, p 14.24. CDPA 1988, s 1(1).25. Horn, above n 13, p 14.26. L Bently ‘Authorship of popular music in UK copyright law’ (2009) 12 Information,Communication and Society 179 at 192.27. R Middleton ‘Work-in(g)-practice: configurations of the popular music intertext’ inTalbot, above n 13, p 77.28. W Fisher Promises to Keep: Technology, Law, and the Future of Entertainment (Stanford,CA: Stanford Law and Politics, 2004).29. L Lessig Free Culture: How Big Media Uses Technology and the Law to Lock DownCulture and Control Creativity (New York: Penguin Press, 2004) ch 10.



more collaborative rap music – even though it had a long and detailed history.30

Likewise, the music industry has been very slow to adapt to the development and riseof online music, particularly with regard to sampling and remixing.

If the law were to take a more ontologically correct stance, it would enable themusic industry to exploit new forms of distribution rather than having to rely onontologically unsound property paradigms. It is not that the music industry automati-cally dislikes adapting itself to new technology – it is because the ontological structureof the law is a property-based one, which is not so applicable for newer technologiesor musical formats. Landes and Posner,31 in their attempts to establish the optimallevel of copyright protection, have suggested that protection should not involve toomany costs for re-users of copyright content, nor that protection should be so light soas to dissuade future investment.32 What was not identified, however, was how legalontology related to their study.

THE ONTOLOGY

For law to hold the ‘correct’ ontology for copyright works (such as music) wouldincrease the likelihood that the legal system would permit, and encourage, optimallevels of exploitation – regardless of technical change. This raises the issue of what the‘correct’ ontology should be. A law that has a correct ontology would be one whichresponds without unexpected side effects, such as unexpected injustice or unintendedconsequences.33 The question of how to build up a correct ontology in relation tomusic has often led to debate as to how to classify the ways in which different typesof music are created. Thus, Davies in Musical Works and Performances34 dividesmusic into five distinct types. The divisions, which he refers to as a ‘slice [of] thecontinuum’,35 separate out types of music based on ‘places where there is change inthe criteria on which we base our understanding and evaluation of what is happen-ing’.36 Likewise, Horn has devised nine categories by which a listener may classifymusic.37 However, such an ontological construction merely seeks to place proprietarylabels over certain genres of works. A correct ontology of musical work would focuson how the music itself is created and the stages involved.

The main concern with copyright law is its ready acceptance of the notion of theromantic author.38 This has obscured how many copyright works are made, and therole that earlier works play in this.39 In relation to music, Straw makes the point thatmusical creation is often more devolved than many other types of copyright works –

30. Brackett, above n 22, p 137.31. W Posner and W Landes ‘An economic analysis of copyright law’ (1989) 18 Journal ofLegal Studies 325.32. Ibid, at 332.33. DR Koepsell The Ontology of Cyberspace: Philosophy, Law, and the Future of Intellec-tual Property (Chicago IL: Open Court, 2000).34. S Davies Musical Works and Performances: A Philosophical Exploration (Oxford:Clarendon Press, 2001) p 9.35. Ibid.36. Ibid.37. Horn, above n 13, p 18.38. J Griffin ‘The evolution of authorship under technology: implications for copyright law’[2005] Intellectual Property Quarterly 135.39. Ibid, at 140–142.



like cinema, he suggests, ‘the problem of isolating authorship within it is no moreeasily resolved’.40 Paintings and novels lend themselves more to the concept of thesingle, inventive romantic author than music does. They are not so overtly collabora-tive in nature.

Middleton implicitly supports Barthes’ work ‘The death of an author’,41 particu-larly in relation to synthesis and electro music.42 Straw suggests that practices such assampling mark the end of authorship,43 in conjunction with comments by musicianBrian Eno who has suggested that there is a need to shift focus away from the conceptof a single ‘genius’ to a group ‘scenius’.44 He believes that individual collaborationsare ‘minor and transitory’.45 Case-law is also littered with situations where one authorseeks to apportion royalties, and cases such as Hadley46 dealing with Spandau Ballet’smusic merely serve to emphasise how complex the relationships between the variousartists can be.47

Whilst these views indicate that music is a highly collaborative endeavour, there isno clear answer as to how to build up an ontology that explicitly takes into accounthow music is created. However, the aforementioned analysis does suggest that thecomplex issue of collaboration needs to be assessed on a case-by-case basis. Therewould not be an automatic allocation of proprietary rights, but rather a continualassessment of how one work is utilised in a later work.

THE CONSEQUENCE OF AN INCORRECT ONTOLOGY

The issue with the current copyright system is that it does not sufficiently take intoaccount collaborative creative endeavour. Whilst copyright law does have rules that setout how to deal with collaborative works, the issue is that these rules are still basedaround proprietary notions. The cases reveal the difficulty in establishing proprietaryboundaries of musical works.

The consequence of the emphasis of the law upon ownership and control overaspects of musical works reveals itself in the use of samples by musicians. Forinstance, it is generally believed that music should consist of short and not longsamples.48 Certain types of copyright are fought for more strongly than others –‘Because contributions to music and lyrics are rewarded with longest and strongestrights, with corresponding financial flows, the division prompts contests betweencontributors to the sound artefact who created those aspects’.49 The case of William-son50 also identifies an issue with the test for infringement of musical copyright. Once

40. Straw, above n 21, p 1.41. R Barthes ‘Death of an author’ in S Heath (ed) Image, Music, Text Essays Selected andTranslated by Stephen Heath (New York: Hill and Wang, 1977).42. Middleton, above n 27, p 63.43. Straw, above n 21, p 206.44. Ibid.45. Ibid.46. Hadley v Kemp [1999] EMLR 589.47. R Arnold ‘Case comment: are performers authors?’ [1999] European Intellectual PropertyReview 464.48. Eg ‘the prevalent assumption that sampling a 35 second saxophone solo would infringecopyright in a musical work of six minutes length’: Bently, above n 26, at 192.49. Ibid, at 195.50. Williamson v Pearson [1987] FSR 97.



the taking of a substantial part has been established, the issue of similarity is assessedin relation to the two final works. The test is what is heard by the ear, and not anote-by-note comparison.51 ‘As Mr Pumfrey said, you look at it as a whole and ask nothow the effect is produced, but what is the effect’.52 Whist this helps to identifyparticular themes, it does not help to detail the manner in which the music was created.For instance, it might be possible for a later piece of music to be substantially basedon an earlier piece, yet not sound similar to the ear. This is particularly relevant in anera where pieces of music can be easily digitally edited and adapted.

The failure to acknowledge sufficiently the manner in which musical works arecreated has led to accusations that the law reinforces old methods of distribution, ratherthan encouraging new methods. This has led to widespread breaches of copyright lawin the field of p2p (peer-to-peer) networking. There has been an increasing trend forcontent recipients to take copyrighted music and to edit and remix it, and then to uploadthese edited versions via p2p networks. As a result of this, those services that permit theuploading of such music to others have been found liable under the rules relating tosecondary liability, or a breach of the right of authorisation. For instance, in the UnitedStates, services such as Napster53 have been found liable of secondary liability, andGrokster54 was found liable for inducing copyright infringement. Although there are noUK cases on the issue, inAustralia, Sharman Networks were found liable for infringingthe right of authorisation.55 In addition, those who use these services to downloadinfringing files can be identified through court orders56 and others have been sued.57

Nonetheless, online infringement continues in large numbers – and of musical copy-right it has been said that ‘Youth culture has forced the hand of tradition, it’s forcedaristocracy to retreat. Aristocracy says. “You can’t do that with these machines”, butyouth culture finds a loophole. There are no rules! The barbarians are taking over’.58

Putting to one side for a moment the issue of law enforcement, such a statement isthe consequence of a mismatched ontology. The subsistence of, and infringement of,musical copyrights obscures the manner of creation with proprietary conceptualisa-tions. It is not the case, per Middleton, that there might need to be changes in societyat large to permit a ‘broader re-evaluation of “individualism” and “ownership” ’.59 Itis simply that the law needs to adapt to an ontologically correct state.

Part of the problem stems from the narrow judicial focus on the argument of theparties and not on the wider purposes of copyright and narrow discussion of legal ruleswithout considering the broader issues which such discussion raises. Cases whichhave looked at joint authorship are particularly instructive. Discussion that concerns

51. Ibid, at 108.52. Ibid, at 111.53. A&M Records, Inc v Napster, Inc 239 F.3d 1004 (9th Circuit, 2001).54. Metro-Golden-Meyer Studios, Inc v Grokster Ltd 125 S.Ct. 2764 (Supreme Court, 2005).55. Universal Music Australia Pty Ltd v Sharman License Holdings Ltd [2005] FCA 1242.56. For the UK, see the website available at http://www.theregister.co.uk/2008/08/20/davenport_lyons_25000/. In the US see Re Verizon Internet Services, Inc 240 F. Supp. 2d 24, 26(DDC, 2003); Re Verizon Internet Services, Inc, 257 F. Supp. 2d 244, 246–247 (DDC, 2003);Recording Industry Association of America, Inc v Verizon Internet Services, Inc 351 F.3d 1229(DC Circuit, 2003). See, inter alia, A Kao ‘RIAA v Verizon: applying the subpoena provision ofthe DMCA’, (2004) 19 Berkeley Technology Law Journal 405.57. See, inter alia, the website available at http://www.theregister.co.uk/2010/01/05/tenenbaum_files_for_retrial/.58. Middleton, above n 27, p 79 quoting The Guardian 23 January 1998 at 2–3.59. Ibid, p 86.



originality in joint authorship, for instance, revolves around the degree of originalityrequired. Godfrey v Lees60 indicates that less originality is required for establishingjoint authorship than there is for copyright subsistence. Should this difference exist?The property paradigm is not ontologically correct. Authorship, notwithstanding themodel of romantic authorship, has always involved some re-use of what has gonebefore. The boundary of originality has always remained low rather than beingabolished, because, to do otherwise, would accept the invalidity of the romantic modelof authorship.

A similar level of artificiality lies around those cases that look at the relevantcontributions of the various parties involved. At first glance, the discussion in thesecases appears to be detailing precisely the steps involved in creating a particular pieceof music. In Hayes,61 for instance, there is a detailed discussion going through theinitial stages of creation, namely how one artist contributed some of the lyrics forsome versions of the song and not others; in Bamgboye62 the manner in which theBlueBells music was influenced by the violin is discussed in considerable detail.However, whilst that discussion has merit, as indicated earlier, artists prefer to try toestablish control over certain copyrights in order to protect their revenue flow. Thisskews the discussion somewhat.63

Copyright infringement cases, such as Sawkins,64 have extremely limited detail inrelation to how earlier works are used in later works. Once copyright subsistence isestablished over a particular ‘work’, the question is whether there has been taking ofa substantial part. The next test is then to ask whether there is substantial similarity.65

Whether or not this paradigmatic partitioning of works in parcels of property reflectsthe creative process is overlooked – and this is in stark contrast to those cases of jointauthorship where the creative process is at least considered.

It has been suggested by Bently that the issue of authors competing over differenttypes of copyrights could be overcome by the creation of another proprietary copy-right over sounds.66 However, this would not resolve the wider issue that the law, increating a property paradigm, has been led down an avenue which no longer accu-rately maps onto the manner in which creative music (or works in general) areproduced.

This naturally raises the question of what an appropriate ontology would be, andthen the question of how this ontology would be applied in practice. It will involve amove away from property conceptions and require the development of a system whichemphasises the creative processes. The move need not be so radical. It merely requiresthe courts to consider in more depth the creative processes involved in a work, and amanner through which creativity can be evaluated and valued.

THE NEW APPROACH

The Statute of Anne embodied the notion of the romantic author, an individual whocreated works in isolation without recourse to earlier works. Some musicologists have

60. Godfrey v Lees [1995] EMLR 307.61. Hayes v Phonogram, above n 14.62. Bamgboye v Reed [2004] EMLR 5.63. Bently, above n 26, at 195.64. Sawkins v Hyperion, above n 2.65. Designers Guild v Williams, above n 17.66. Bently, above n 26.



suggested that this is consistent with the notion of classical music, where a composerwould be creating his works as a single genius. By contrast, ‘copyright law doesn’tmesh with the practice of popular music’.67 It has been suggested that modern popularmusic is more of a communal work, which is less reliant on the contribution ofparticular individuals. To quote Straw, ‘. . . we might consider Brian Eno’s claim thatcreativity now operates at the macro-level of whole genres, not at the micro-levels ofindividual artists. In place of “genius”, Eno awkwardly suggests, we must speak of“scenius” ’.68 If this view is accepted, then it follows that, for the law to be correctlyontologically founded, then there should be greater consideration of how earlier worksare re-used. However, the notion of the romantic author developed as a device toidentify authors, and to make those authors’ works more valued as an asset. To acceptthat romantic authorship is of less relevance, as in popular music, endangers the useof the concept of authorship and copyright to enhance the value of assets.69 Theproperty paradigm of copyright, consequently, is less likely to be a paradigm that willencourage deeper consideration of the steps involved in the making of music.

The bluntness of a property-based notion of copyright is demonstrated in theapproach of the recording industry to remixes. For instance, rather than seeking toencourage the re-use of their works, they would prefer to utilise copyright law to eitherrestrict re-use or increase revenue income on specific samples.70 A system of copyrightwhich moves away from the property paradigm could be one which would lead tomore emphasis on the origin and manner of making a work or piece of music.

The development of copyright as a propertised right was not a given. One historicalbranch of copyright cases which dealt with abridgements and translations emphasisedthe ‘merit’ of a potentially infringing work. In Gyles v Wilcox,71 Lord ChancellorHardwicke, having briefly referred to the stated aims of the Statute of Anne, stated thatthe ‘. . . Act is not to be construed strictly, but according to the Intention of theLegislature: However, the intention must be formed from the words of the Statute’.72

What this resulted in was a specific shift of focus – the question became ‘Whether thesecond Book has been the same Book with the former?’,73 though, in this case, itboiled down to a specific question of whether the abridgement was ‘a Work ofJudgment’.74

There were also judicial attacks on the use of theoretical property principles beingapplied to intellectual works, these criticisms being led by Willes J in Millar vTaylor:75

‘Metaphysical reasoning is too subtile; and arguments from the supposedmodes of acquiring the property of acorns, or a vacant piece of ground in animaginary state of nature, are too remote. Besides, the comparison does not holdbetween things which have a physical existence, and incorporeal rights.’76

67. Toyrbee (2004), p 127, quoted in ibid, at 180.68. Straw, above n 21, p 206.69. Ibid.70. Lessig, above n 29, p 285.71. Gyles v Wilcox Barn C 368 (1741).72. Ibid, at 368.73. Ibid, at 369.74. Ibid.75. Millar v Taylor (1768), [1558–1774] All ER Rep 119.76. Ibid, at 218.



However, the system that ultimately prevailed was the one based around, and whichencouraged, a property-based way of thinking. The ‘merit’ approach had becomemoribund by the Copyright Act 1911. Nonetheless, in the 1990s, a defence of reverseengineering for digital works developed, where, once again, the property paradigmwas dispatched in favour of a system espousing the details of how earlier works arere-used in later works.

A system that develops those principles of re-use, and moulds them into a legalsystem which shifts away from the property paradigm, would embrace a more onto-logically accurate understanding of how works are made and created. To do so wouldlead to a system of regulation that embodies the changes made possible by technology,for the technology is relevant principally in relation to the way in which it is utilisedin re-use.

RESTRICTING THE PROPERTY PARADIGM

The development of the property paradigm of copyright has been encouraged throughthe notion of the romantic, individualist author. This would suggest that a startingpoint for a restricted system would be to focus upon how an author (or composer)utilises their surroundings, and possibly earlier works (or musical pieces) to make‘new’ works.

The works of Davies, Straw and Middleton et al77 all refer to the fact that earliermusical works play a role in the creation of later musical works. Nonetheless, there islittle clarification as to the mechanisms by which this is achieved. One would alsoexpect works that consider the wider notion of ‘creativity’ to make some reference tothe mechanisms of how earlier works are influential. However, they focus on the useof knowledge after it is obtained.78 One exception is the work of Csikszentmihalyi,79

but even his work does not detail the actual process of how information in one workis abstracted and then re-used.

An approach which does detail how existing knowledge and works are re-used canbe found through study of the historical judgments which favour focusing on the meritof the potentially infringing work.80 It can also be found by considering the cases thatconcern digital reverse engineering. Both approaches make reference to how an earlierwork may be re-used as part of the process in creating a subsequent work. Bothapproaches eschew property paradigm thinking.

The approach of the earlier cases can be seen to mirror closely some of the workof philosopher John Locke. This is particularly true of Lord Camden’s judgment inDonaldson v Beckett,81 which, although not citing Locke as his fellow judges do in

77. Straw, above n 21; Middleton, above n 27, and Davies, above n 34.78. J Guilford ‘Creativity’ (1950) 5 American Psychologist 444; R Weisberg Creativity:Beyond the Myths of Genius (NewYork: Freeman, 1993) p 242; M Csikszentmihalyi Creativity:Flow and Psychology of Discovery and Invention (New York: Harper Perennial, 1996)pp 77–106 (in light of 127–147), 182 and 342; B Hennessey and T Amabile ‘The Conditions ofCreativity’ in R Sternberg (ed) The Nature of Creativity (Cambridge: Cambridge UniversityPress, 1998) pp 11–38, particularly 29–30 on rewards. For a collection, see M Runco andR Albert Theories of Creativity (London: Sage, 1990).79. Csikszentmihalyi, above n 78.80. See the previous section.81. The Hansard Report of Donaldson v Beckett, reported as ‘Proceedings in the Lords on theQuestion of Literary Property, 14 Geo III 1st Ser 17 950 (1774).



relation to Locke’s Second Treatise,82 closely mirrors Locke’s work in the Essay onHuman Understanding.83 Locke argued ideas arise from the combination of existingthoughts. These thoughts arise from the observation of events.84 A thought can becreated that is greater than the combination of its parts.

Locke details the process by which individuals may perceive and understand theoperation of things around them. Locke gives an example concerning gold. He statesthat we know that fire can melt gold, and, likewise, from this we know that gold maybe melted.85 We thus learn from our surroundings by observation, and we then applywhat we learn when we perform an act in the future.86 Locke develops this further byreference to ‘passive’ and ‘active’ powers,87 ‘passive’ powers being those such asobservation and listening, and ‘active’ powers being the re-use of the informationgleaned. Locke provides an example of billiards, where if a ball is hit with a cue, an‘active’ power is at work moving the ball.88 The observer may, in turn, learn how tocarry out that ‘active’ power.89 To apply this to music, someone listening to it may‘passively’ listen to the music, but then ‘actively’ re-use what has been learnt in eithermaking a new piece of music, or in reproducing that music using the knowledgegained. That knowledge which is reproduced may contain copyright elements. An‘active’ power may thus result in an infringement of copyright. However, Locke’snotion of powers is only intended as a description of how objects relate to one another– Locke is not advocating that all ‘active’ powers be legally possible or desirable.90

Lockean conceptions of knowledge are essentially representing that which isknown today as reverse engineering – ‘starting with a known product and workingbackward to divine the process which aided its development or manufacture’.91 In thecontext of modern day copyright, reverse engineering is a term exclusively used inrelation to digital technology. However, reverse engineering is not a new concept.92

Reverse engineering is a process that individuals use in everyday thought. Whenseeking solutions to problems, individuals will ‘deconstruct’ them, in order to con-struct an appropriate solution. There are a number of examples. For instance, amedical doctor will research how a particular medicine works before administering itto a patient. The same is true of mechanics solving problems with cars, and chefsassessing what a recipe might be for a type of food.93 ‘What it means, broadly

82. J Locke, The Second Treatise of Government and A Letter Concerning Toleration(reprinted in New York: Dover Press, 2002).83. J Locke Essays on Human Understanding (1689) (Leeds: William Tegg & Co, 1880).84. Ibid, Book IV, ch I at §8.85. Ibid, Book II, ch XXI, at §1.86. Ibid.87. Ibid, at §2.88. Ibid, at §4.89. Ibid, at §4.90. Ibid, at §73.91. Kewanee Oil Co v Bicron Corp 416 US 470 (1974) at 476, and EJ Chikofsky and JH CrossII ‘Reverse engineering and design recovery: a taxonomy in IEEE software’ (1990) 7(1) IEEESoftware 13 at 15 who state that reverse engineering is ‘. . . the process of analyzing a subjectsystem to identify the system’s components and their interrelationships and create representa-tions of the system in another form or at a higher level of abstraction’.92. P Samuelson and S Scotchmer ‘The law and economics of reverse engineering’ (200) 111Yale Law Journal 1575 at 1577.93. A Johnson-Laird ‘Reverse engineering of software: separating legal mythology fromactual technology’ (1992) 6 Software Law Journal 331 at 334.



speaking, is the process of extracting know-how or knowledge from a human madeartefact.’94 In the context of music, it has been described as part of the learningprocess:

‘it’s my role to model how to learn and practice each piece of music. Just likeworking through a reverse engineering problem, my son and I first break each piecedown into its individual phrases. We then experiment with different strategies forlearning each phrase, often further breaking a phrase down into smaller parts. Wefinally “solve” a piece by slowing putting its parts and phrases back together.’95

As a result of this, it is suggested that any reformed copyright system that acceptsthe notion of re-use and which moves away from the property paradigm should takeinto account how creators such as musicians (a) understand earlier works and (b)re-use the knowledge obtained.

DIGITAL REVERSE ENGINEERING

Reverse engineering has developed as a ‘hot topic’ in the digital sphere because of theadditional information that a digital copyright work, such as a digital recording ofmusic, may hold. In relation to music, digital reverse engineering can ‘help to quicklyidentify, select and isolate interesting features in a mix, such as individual notes,instruments and voice’.96 The difference between reverse engineering a non-digitalwork and a digital work is apparent when considering musical works. For example, if,in the non-digital context, a listener wishes to ‘reverse engineer’ a piece of music, thelistener can study the piece closely and then, perhaps, imitate it or make a transcriptionof it. Although a listener may learn a lot from this, it does not reveal the underlyingprocesses used. ‘Hidden’ information, such as the specific types of instruments used,can only be guessed at through listening to the music.

In contrast, the reverse engineering of digital music may reveal ‘hidden’ infor-mation. This will depend on the type of file used to save the music. If it is merelyan MP3 recording, there is only the final product, as with a standard analoguerecording. However, if it is music saved in a rich computer format,97 then therecould be a lot of information available. Above and beyond the information gleanedfrom merely listening to the music, it may be possible to load the music into theoriginal program that created it, or in which it was composed. This will revealprecise information as to the instruments used, tempo and so on. If it is not possibleto load the music into a program, then a reverse engineer could look at the code thatmakes up the music. By looking at the hex decimal code, it may be possible toidentify the program used to create the work, and it could reveal what sort ofplug-ins might be running. To the skilled user, it might also reveal how theseplug-ins operate. Reverse engineering reveals most information when the music is

94. Samuelson and Scotchmer, above n 92.95. See the website available at http://www.wired.com/geekdad/2007/07/looking-to-help/.96. See the website available at http://www.sonicstate.com/news/2009/04/01/messe09-reverse-engineer-your-music/.97. A rich format is one that contains identifiable commands. For detailed discussion seeT Simcoe Open Standards and Intellectual Property Rights, available at http://www.rotman.utoronto.ca/timothy.simcoe/papers/OpenStandards_IPR.pdf.



created within software engines,98 because the skilled user can identify any addi-tional code which has been used for specific effects.

REVERSE ENGINEERING PRINCIPLES IN COPYRIGHT CASE-LAW

The failure of UK courts to consider adequately the process of reverse engineering isclear in cases dealing with copyright infringement. One of the key cases is Sawkins,particularly in regard to the reasoning employed by Mummery LJ. As was noted at thestart of the paper, in relation to copyright infringement in music he states that, ‘Thetest of substantial reproduction is not a note-by-note textual comparison of the scores.It involves listening to and comparing the sounds of the copyright work and of theinfringing work’.99

Why should similarity of sound play a role (and visual similarity if appropriate)?100

A work that sounds dissimilar could still have made as much use of an earlier work –and the flaw of this can be seen in earlier cases such as Hawkes and Son,101 where anamount taken may be so small that it may be impossible to recognise the earlier work.

There were particular complaints addressed against some further dicta ofMummery LJ, for, in a subsequent paragraph, he argued that:

‘It is wrong in principle to single out the notes as uniquely significant forcopyright purposes and to proceed to deny copyright to the other elements thatmake some contribution to the sound of the music when performed, such asperforming indications, tempo and performance practice indicators, if they are theproduct of a person’s effort, skill and time, bearing in mind, of course, the “rela-tively modest” level of the threshold for a work to qualify for protection.’102

By permitting copyright over elements such as performing indicators and tempo, ithas been suggested that copyright protection has extended to a level that is not takinginto account the consequences upon future musical works.103 This has been a criticismparticularly aimed at popular music, especially in relation to sampling. This contrastsstrongly with those cases which are not focusing on infringement but on issues of jointauthorship. In those cases, there is much debate as to the processes used in the creationof a piece of, say, music. However, in relation to infringement cases, the discussion ismuch less developed. In Sawkins, there is discussion of how Sawkins took the originalpieces by Lalande and mixed in his own labour, skill and effort, but the argument ismore focused on establishing copyright in the plaintiff’s work rather than in the widerprocesses of creation. Furthermore, somewhat critically, the issue of subsistence isbeing decided in relation to the plaintiff’s work and not that of the defendants.

Notwithstanding this lack of consideration, there are various principles throughwhich courts may permit certain non-infringing acts. However, these do notsufficiently consider how and why authors and musicians may re-use earlier works.

98. For example, the Quake 3 engine called idTech3; see the website available at http://www.moddb.com/engines/id-tech-3.99. Sawkins v Hyperion, above n 2, at 3295.100. D’Almaine v Boosey (1835) 1 Y & C Ex 288; Austin v Columbia Gramophone Co Ltd[1917–23] MacCC 398; Francis Day & Hunter Ltd v Bron [1963] Ch 587.101. Hawkes & Son (London) Ltd v Paramount Film Service Ltd [1934] Ch 593.102. Sawkins v Hyperion, above n 2, at 3295.103. See above n 4.



Insufficient originality in the original work remains a concept which focuses on theplaintiff’s and not the defendant’s work.104 The same is true of the concept of thenon-protection of ideas – focus is not on how these are re-used, but on what has beentaken from the plaintiff’s work.105 Independent creation, for instance, will not merit afinding of infringement, but this is essentially irrelevant where there has beenre-use.106

In order to ensure that courts consider how creative re-use may be being affectedin any given copyright dispute, every court should be required to ask itself two mainquestions:

1. What reverse engineering has taken place?2. Will it be impeded by the decision?

However, the degree to which reverse engineering should be allowed has not yetbeen discussed. If there were no limits, it would leave the issue of infringement moot.Consequently, we will now proceed to consider how far the law has permitted reverseengineering and what its limits are. Discussion of reverse engineering in the UK andUSA is limited to the digital environment, but we will draw out principles which maybe considered within the context of analogue works.

THE LAW ON DIGITAL REVERSE ENGINEERING – THE UK APPROACH

The UK law concerning digital reverse engineering encourages some consideration ofthe processes involved in making a copyright work, such as that involved in writing apiece of music. However, the focus of the law is not on works such as music per se –it is instead on decompilation, principally of computer software. Decompilationdescribes the process whereby code of low level abstraction is turned into a higherlevel of abstraction, in order to make it readable for humans. Decompilation is not aninfringement of UK copyright under s 50B of the CDPA 1988 if:

‘(2) (a) it is necessary to decompile the program to obtain the informationnecessary to create an independent program which can be operated with theprogram decompiled, or with another program (the “permitted objective”), and

(b) the information obtained is not used for any purpose other than thepermitted objective.’107

The decompilation provision, s 50B, is relatively broad in that it permits decom-pilation for the purposes of interoperability with another computer program other thanthat which is decompiled. Thus, it would be possible to decompile a file format for use

104. Davies et al, above n 7, at 3–128.105. Ibid, at 2–06.106. X and G. Ricordi & Co, etc Ltd v Clayton and Waller Ltd [1928–35] MacCC 154.107. CDPA 1988, s 50B. The section is implementing Directive 91/250/EC on the LegalProtection of Computer Software, OJ L122/42. This section should be read in conjunction withs 50BA. That section permits certain acts for the observation, study and testing of computerprograms: ‘(1) It is not an infringement of copyright for a lawful user of a copy of a computerprogram to observe, study or test the functioning of the program in order to determine the ideasand principles which underlie any element of the program if he does so while performing anyof the acts of loading, displaying, running, transmitting or storing the program which he isentitled to do’.



in another program, or to decompile one program in order to obtain interoperabilitywith another. In opening up a program in this manner, a re-user can more easilyunderstand how the work operates. This is consistent with encouraging certain re-uses,because content recipients will be able to obtain more access to content. It thus hassome general applicability in terms of musical copyright, although this is clearlysomewhat abstract. What this demonstrates more is that, for the reverse engineeringapproach to succeed in relation to music, it must have more general principles ratherthan such specific rules. For this reason, we now turn to US law.

US REVERSE ENGINEERING CASE LAW UNDER §107

US law has developed around the general fair use provision in the US Code108 and themore recent equivalent provision under the Digital Millennium Copyright Act(DMCA).109 The principles that have developed under US law are of more directrelevance in the field of musical copyright. We will begin by assessing case-law basedaround the US Code. In the Ninth Circuit case of Sega v Accolade,110 Accolade hadreproduced elements of Sega’s code in order that their computer games could beplayed on the Sega ‘Genesis’ games console. Near the start of the judgment, CircuitJudge Reinhardt wrote that:

‘Accolade used a two-step process to render its video games compatible withthe Genesis console. First, it “reverse engineered” Sega’s video game programs inorder to discover the requirements for compatibility with the Genesis console. Aspart of the reverse engineering process, Accolade transformed the machine-readable object code contained in commercially available copies of Sega’s gamecartridges into human-readable source code using a process called “disassembly”or “decompilation”.’111

Lexmark v Static Control112 and Atari v Nintendo113 follow a similar route – ofassessing the notion of reverse engineering toward the start of the judgment,114 andthen the reason for performing the reverse engineering. They consider how the reverseengineering was achieved and what steps were involved, and then how the knowledgegained was used to create a subsequent work.

The manner here in which the court focuses on how earlier works are re-used inlater works has clear applicability to the re-use of musical works. As noted earlier inthe paper, in many copyright infringement cases concerning music there has beenlimited attention paid to the creative processes involved. Actions for copyrightinfringement focus on the final musical products of reverse engineering, whereascases of digital reverse engineering are looking at the creative process.

108. 17 USC §107.109. The Digital Millennium Copyright Act, Pub. 105–304, 28 October 1998, 112 Stat 2860(DMCA).110. Sega v Accolade 977 F.2d 1510 (9th Circuit, 1992).111. Ibid, at 1514.112. Lexmark v Static Control 387 F.3d 522 (6th Circuit, 2004).113. Atari Games Corp v Nintendo of America Inc 975 F.2d 832 (Federal Circuit, 1992).114. Lexmark v Static Control, above n 112, at 529; Atari v Nintendo, ibid, at 836.



THE US TEST OF ‘NECESSITY’

When individuals do manage to reverse engineer a work such as a piece of music, theyhave to be extremely careful to re-use only the knowledge necessary. In Nintendo vAtari, Judge Reader of the US Court of Appeals, Second Circuit, wrote that, inaddition, the grounds of breaching Copyright Office rules, an action by Nintendoagainst Atari was possible because:

‘Fair use did not give Atari more than the right to understand the 10 NESprogram and to distinguish the protected from the unprotected elements of the 10NES program. Any copying beyond that necessary to understand the 10 NESprogram was an infringement.’115

According to Judge Reader, the test of necessity116 in Nintendo was adapted fromComputer Associates v Altai117 (another Second Circuit case) where the court was tofilter out as unprotectable the ideas, unoriginal expression, expression incident to theidea and expression already in the public domain.118‘Necessity’ is a potentially narrowtest, limiting the scope of any reverse engineering – thus, an individual reverseengineering a protected (ie encrypted) piece of music should only do so as far asnecessary to gain access to unprotected elements. However, Circuit Judge Reinhardtin Sega v Accolade119 maintained a more restrictive approach towards reverse engi-neering, by referring to the 1879 case of Baker v Selden.120 Circuit Judge Reinhardtcited a part of Baker which considered the reproduction of elements of a house-keeping text. Justice Bradley had stated that:

‘Where the art it teaches cannot be used without employing the methods anddiagrams used to illustrate the book, or such as are similar to them, such methodsand diagrams are to be considered as necessary incidents to the art, and giventherewith to the public; not given for the purpose of publication in other worksexplanatory of the art, but for the purpose of practical application.’121

By referring to Baker in this way, Circuit Judge Reinhardt highlighted that the‘necessity’ test should be treated as meaning ‘necessary incidents to the art’.122 In themain text of his judgment he went further:

‘The unprotected aspects of most functional works are readily accessible tothe human eye. The systems described in accounting textbooks or the basic struc-tural concepts embodied in architectural plans, to give two examples, can be easilycopied without also copying any of the protected, expressive aspects of originalworks. Computer programs, however, are typically distributed for public use inobject code form . . . [which ordinarily cannot be read by humans].’123

115. Atari v Nintendo, ibid, at 844.116. Note that in the UK CDPA 1988, s 50B uses the word ‘necessary’, but there is no case-lawon its interpretation.117. Computer Associates v Altai 982 F.2d 693 (2nd Circuit, 1992).118. Atari v Nintendo, above n 113, at 839.119. Sega v Accolade, above n 110, at 1524.120. Baker v Selden 101 US 99 (1879).121. Ibid, at 103, cited by Sega v Accolade, above n 110, at 1524 but in reference to Baker vSelden at 104 (103 introduces the argument at 104).122. Sega v Accolade, ibid, at 1524123. Ibid, at 1525.



The implication from this is that digital works should be capable of being reverseengineered up to the level of analogue works. However, digital reverse engineeringmay go further than this. Code may contain more information, for instance, as to howa piece of music was constructed, or particular intimate details, and these may not becopyrightable. Consequently, a test is suggested that should require the courts toconsider whether reverse engineering is being impeded and then whether it is beingimpeded to an undesirable degree. The first two parts of this test were highlightedearlier:


However, an additional third part would ensure that these provisions are kept withinsome discretionary limits:

3. Has reverse engineering been permitted so far as is necessary for the re-use in relationto non-copyright elements?

By having a test that should be followed by the courts in any case of copyrightinfringement, it is hoped that sufficient regard will be given to the factors thatinfluence the creative re-use of copyright content such as music. Such a test is notwithout possible issues, most notably that certain re-uses may become favoured.

DOES REVERSE ENGINEERING FAVOUR CERTAIN RE-USERS OFCOPYRIGHTED MUSIC?

It has been suggested that reverse engineering provides a sound basis upon which anunderstanding of the re-use of copyright works such as music could be based. Thisbasis may then be used for reform of the law when considering how right holderscould be compensated for the re-use of their works. Reverse engineering laws avoidthe shortcomings of many aspects of the property paradigm. They help to encouragethe courts to focus upon the creative processes involved in making works, rather thanto focus too heavily upon existing financial interests of right holders.

The growth in digital technology has meant that there is increasingly the possibilityto reverse engineer digital music. However, dominant right holders, because of theirresources, can develop digital rights management (DRM) mechanisms at a sufficientpace to make circumstances difficult. For instance, Microsoft invested considerablemoney into the DRM mechanisms for Windows Media Player (which can concernmusical works) and used various obfuscation techniques to disguise how it worked. Asnoted by the hacker ‘Beale Screamer’ who partly reverse engineered the system in2001, these were complex.124 At present, there are still (!) no programs available thatcan fully imitate that Media Player DRM system.125

The application of reverse engineering is not without its issues. The current rulesof reverse engineering favour those who have the resources (or skills) to reverse

124. As at 1 May 2010. Details of the operation of the system are available at http://www.bearcave.com/misl/misl_tech/msdrm/readme.html.125. Ibid. The closest nearest success is detailed at http://www.mydigitallife.info/2006/09/09/crack-remove-and-disable-windows-media-player-drm-license-acquisition-and-music-copy-protection-with-fairuse4wm/ but this only works on content that is already licensed.



engineer DRM mechanisms.126 The system places at an advantage those who haveskills related to the breaking of mechanisms designed to defeat reverse engineering,rather than those whose skills may be concerned with other creative re-uses ofcopyright music. Those who cannot circumvent DRM mechanisms are at a disadvan-tage. This means that more technical uses are favoured. An additional issue is that thelegal conception of reverse engineering tends to favour those coders who have theresources (or skills) to develop more complicated mechanisms. It was noted earlier,for instance, that Microsoft has only had its DRM in Windows Media Player partiallyhacked.

To deal with these issues, the proposed test of necessity, when considered by acourt, should be broken down into two component stages. The first is the stage atwhich the DRM mechanism is circumvented. The second stage is the actual analysisand re-use of content. By way of example, a reverse engineer of a computer programneeds skills to be able to access compiled source code. The reverse engineer can thenmake this code human readable, which will involve de-compilation of the object code.Once this is done, different skills can be employed to understand that human readablecode and this may be done by other individuals. Any application of the test ofnecessity therefore needs to bear in mind that access to aspects of musical works maybe dependent on circumventing the DRM mechanism, which may be difficult to break.

RECENT US DMCA ACCESS ISSUES

Reverse engineering of music may arise wherever an individual is listening to music.However, in the digital context, in order to gain access to hidden digital information,it may be necessary to break an access mechanism. When access is granted to a workit is now invariably protected by a licence and a DRM mechanism.

The DMCA has limited the applicability of the reverse engineering defence whena DRM mechanism is involved. The issue is important in relation to music becausemusic is often protected by DRM mechanisms. Reverse engineering is then governedby §1201(f):

‘(f) REVERSE ENGINEERING – (1) Notwithstanding the provisions ofsubsection (a)(1) (A), a person who has lawfully obtained the right to use a copy ofa computer program may circumvent a technological measure that effectivelycontrols access to a particular portion of that program for the sole purpose ofidentifying and analyzing those elements of the program that are necessary toachieve interoperability of an independently created computer program with otherprograms, and that have not previously been readily available to the person engag-ing in the circumvention, to the extent any such acts of identification and analysisdo not constitute infringement under this title.’127

126. J Griffin ‘The rise of the digital technology meritocracy: legal rules and their impact’(2006) 15 Information and Communications Technology Law 211.127. Additional parts are: ‘(3) The information acquired through the acts permitted underparagraph (1), and the means permitted under paragraph (2), may be made available to othersif the person referred to in paragraph (1) or (2), as the case may be, provides such informationor means solely for the purpose of enabling interoperability of an independently createdcomputer program with other programs, and to the extent that doing so does not constituteinfringement under this title or violate applicable law other than this section. (4) For purposesof this subsection, the term “interoperability” means the ability of computer programs to



The importance of §1201(f) is that it permits only program-to-program interoper-ability, which is much narrower than that of the pre-DMCA case-law just discussed,and it therefore renders the applicability of reverse engineering to music somewhatmoot. It would limit the applicability of the proposed ‘necessity’ test. Nonetheless, thescope of §1201(f) is somewhat complicated by §1201(c)(1) which states:

‘(c) OTHER RIGHTS, ETC., NOT AFFECTED – (1) Nothing in this sectionshall affect rights, remedies, limitations, or defenses to copyright infringement,including fair use, under this title.’

However, it has to be questioned what the point of §1201(c)(1) would be if§1201(f) were strictly followed, for it would render §1201(c)(1) of little use. Subse-quent case-law has indeed placed emphasis on §1201(c)(1). The Sixth Circuit ofAppeals in Lexmark v Static Control stated that the defence would:

‘Apply only when traditional copyright infringement does not occur and onlywhen the challenged actions (in the case of the third provision) would not violateother “applicable law[s]”.’128

In other words, §1201(f) applies primarily when there is a breach of access controlsunder the DMCA. The Sixth Circuit Court of Appeals judgment suggests that if anaction for copyright infringement is alleged, then the pre-DMCA case-law wouldapply. However, the importance of this is limited because a right holder simply neednot attack on grounds of infringement of, say, a musical copyright, but on the groundthat an access control was broken. The stricter DMCA approach would then apply.This is an undesirable position. It means that a party bringing a case can restrict thesituations in which reverse engineering will apply. It is suggested that the DMCAprovision should not apply to any copyright works such as musical works protected byDRM.

CONCLUDING COMMENTS

The current legal system of digital reverse engineering provides many advantages,notably in providing a structure through which courts could consider the re-use ofexisting copyright works, such as music, free from the property paradigm. However,there is the risk that the DMCA could curtail such an approach, particularly as manymusical works are protected by DRM. Any future reforms need to expand consider-ably the scope of the analysis of pre-DMCA reverse engineering case-law.

The application of reverse engineering principles to all instances of re-use ofcopyright music would help at a number of levels. It would ensure that courts wouldconsider the impact of their judgments upon re-use, and mark a move away from theheavy emphasis on the proprietary interests of existing right holders. In implementingthe system, there are a number of questions that courts should be required to ask inorder to resolve any copyright disputes:


exchange information, and of such programs mutually to use the information which has beenexchanged’.128. Lexmark v Static Control, above n 112, at 546.



3. Has reverse engineering been permitted so far as is necessary for the re-use in relationto non-copyright elements?

Parts (1) and (2) are designed only to raise the issues involved, but in relation to (3),the court should not impede reverse engineering that is ‘necessary for a creativere-use’. The proposed system will work in both the analogue and digital context. Forexample, if there is a case of infringement of analogue music (as in Sawkins), the testwill work as follows:

1. What reverse engineering has taken place when the defendant altered the existingmusic?

2. Is the judgment of the court likely to impede this in the present case? This could involveconsideration of the techniques used in producing music, in terms of how content isre-used and also how it is accessed, eg to what degree are p2p networks required as apart of this process?

3. If there is sufficient evidence that the judgment would impede certain re-uses of music(or other works), then these should be permitted so far as necessary for a re-use ofnon-copyright elements.

In the context of a digital copyright case, for instance in relation to DVDs (such asReimerdes)129 rather than music per se, the principles would operate as follows:

1. What reverse engineering was involved in breaking the DRM mechanism?2. Will that be impeded in the present case?3. What re-uses might be impeded? In Reimerdes, musical remixes could have been

considered in terms of the level of copyright protection provided to the samples.

The proposed system is a move away from the current system and its emphasisupon the proprietary interests of right holders. The system will focus much more onthe potential future uses of a copyright work. It is a retreat from the notion of romanticauthorship, to an acceptance of the collaborative notion of authorship. In so doing, theproposed system will be more ontologically accurate and, in turn, creative industriessuch as the music industry will gain a legal system that reflects the creative processesof its creators. As the musician Brian Eno commented at the Sydney LuminousFestival:

‘I was an art student and, like all art students, I was encouraged to believe thatthere were a few great figures like Picasso and Kandinsky, Rembrandt and Giottoand so on who sort-of appeared out of nowhere and produced artistic revolution.

As I looked at art more and more, I discovered that that wasn’t really a truepicture.

What really happened was that there was sometimes very fertile scenesinvolving lots and lots of people – some of them artists, some of them collectors,some of them curators, thinkers, theorists, people who were fashionable and knewwhat the hip things were – all sorts of people who created a kind of ecology oftalent. And out of that ecology arose some wonderful work.’130

129. Universal City Studios, Inc v Reimerdes 82 F.Supp.2d 211 (SDNY, 2000).130. See the website available at http://www.synthtopia.com/content/2009/07/09/brian-eno-on-genius-and-scenius/.



The proposed system will require that courts take into account the nature ofcreativity, and will signal a shift away from the property paradigm that has dominatedcopyright cases. It will be a system that, in Eno’s words, is able to acknowledge andeffectively regulate ‘an ecology of talent’.



copyright in music: a role for the principles of reverse engineering

Documents